Liu, J. (2007). Metadata implementation. In Metadata
and its Applications in the Digital Library: Approaches and Practices (pp.
124-125). Westport: Libraries Unlimited.
Within the Metadata Implementation chapter, there is a small
section on crosswalking that begins with a quote from the Dublin Core Metadata
Glossary that states that “crosswalks help promote interoperability”. The author then moves on to define a
crosswalk as a “high level mapping table for conversion”. Such a method leads to inevitable sacrifice
of data, because it is almost impossible to match all of the elements in the
original schema with those of the other schema.
No two metadata standards are completely equivalent. While this can be useful to pick and choose a
schema which best suits a project, it can also be detrimental when trying to
transfer data from one metadata standard to another. The author mentions several projects that
have used crosswalks extensively, such as Michigan State University’s
Innovative Interfaces XML Harvester, which is a tool to help convert the
collection-level information in EAD into MARC.
METS is described as an example of a standard that avoids transformation
problems due to its acceptance of descriptive metadata in any format regardless
of local rules or standards. In a time
where standards are constantly shifting and advancing, METS is an invaluable
tool.
No comments:
Post a Comment